ggerganov/whisper.cpp

Pull Requests Commits
vulkan: support solve_tri with larger N/K values (llama/17781)
jeffbolznv committed 78 days ago
Verified 875d8614
metal : fix build(#17799)
ggerganov committed 78 days ago
Verified 41cf229d
vulkan: Replace deprecated VK_EXT_validation_features (llama/17637)
rillomas committed 78 days ago
Verified a8d02735
vulkan: Fix mismatch in TOPK_MOE unit test (llama/17541)
rillomas committed 78 days ago
Verified 191e5f46
vulkan: add more num_blocks instantiations in rms_norm (llama/17701)
jeffbolznv committed 78 days ago
Verified 64a3f573
vulkan: fix top_k bug when there are ties in the input (llama/17659)
jeffbolznv committed 78 days ago
Verified 0484147a
vulkan : support conv-2d with large output size (llama/17685)
Acly committed 78 days ago
Verified 0b53759b
ggml webgpu: unary op suppport, code refactoring, ops support (llama/17764)
reeselevine committed 78 days ago
Verified 23984be4
vulkan: enable mmvq for q2_k on NVIDIA (llama/17675)
jeffbolznv committed 78 days ago
Verified 7e97d3b0
vulkan: set all memory allocations to high priority (llama/17624)
jeffbolznv committed 78 days ago
Verified 32ba1ec8
rpc : fix alloc size logic (llama/17116)
ggerganov committed 78 days ago
Verified aefcd75f
metal : add residency sets keep-alive heartbeat (llama/17766)
ggerganov committed 78 days ago
Verified 322903fa
HIP : fix RDNA4 build (llama/17792)
JohannesGaessler committed 78 days ago
Verified 4170159d
Q4/Q8 Tiled Gemm Optimization. (llama/16999)
shalinib-ibm committed 78 days ago
Verified d30b7440
CUDA: fix FA VKQ accumulator overflow (llama/17746)
JohannesGaessler committed 78 days ago
Verified 14502d65
HIP: enable WMMA-MMQ INT kernels for RDNA 3 (llama/17576)
jiachengjason committed 78 days ago
Verified e3f3c6ea
Add support for CUMSUM and TRI for CUDA. (llama/17584)
pwilkin committed 78 days ago
Verified 8d44d618
metal: TRI, FILL, EXPM1, SOFTPLUS (llama/16623)
gabe-l-hart committed 78 days ago
Verified 8902c9d9
ggml-cpu : remove asserts always evaluating to false (llama/17728)
Alcpz committed 78 days ago
Verified f96ebc92
metal : use params per pipeline instance (llama/17739)
ggerganov committed 78 days ago
Verified 194d0164
build : move _WIN32_WINNT definition to headers (llama/17736)
angt committed 78 days ago
Verified 92e50155
ggml-cpu: remove duplicate conditional check 'iid' (llama/17650)
GermanAizek committed 78 days ago
Verified 3794a0d3
CUDA: generalized (mma) FA, add Volta support (llama/17505)
JohannesGaessler committed 78 days ago
Verified 7adbcafb
metal : fix data race in pipeline library (llama/17731)
ggerganov committed 78 days ago
Verified 4a00f2e3
ggml webgpu: add support for emscripten builds (llama/17184)
reeselevine committed 78 days ago
Verified d263bdbf
vulkan: Reduce temporary memory usage for TOP_K (llama/17623)
jeffbolznv committed 78 days ago
Verified 86cb5ab9
cmake : add utf8 compilation options for msvc (llama/17682)
xiaobing318 committed 78 days ago
Verified fffdf679
ggml : use svcntb() for SVE vector length detection (llama/17474)
angt committed 78 days ago
Verified 16688c6d
CANN: Disable Ger operator of OUT_PROD on 310p device (llama/17563)
TianHao324 committed 78 days ago
Verified a64d46a5
ggml : remove redundant n_copies check when setting input/output (llama/17612)
danbev committed 78 days ago
Verified 201b9107
Terms Of Use Privacy Policy Refund Policy Impressum